This shows the incomplete pump cover heater unit, the aluminum block will be cross-drilled with a couple of 3/8 holes, two 5 watt power resisters will then be epoxied inside the holes. I will be ordering the resisters shortly so will post a pic of the finished heater in the future.

pump_cover_heater_-_resized_and_tweeked.JPG (11 Kb, 56 downloads)

This shows a pic of the new 25 gallon fuel tank made from a 100 pound propane tank. This tank sits a couple feet away from the heater inside the house so stays warm once the heater is running but it will cool off enough to thicken the oil if I decide to shut down the heater during the winter. There is a heater well welded across the bottom of the tank. The well is just a piece of 3/4 inch pipe, the center section of this pipe is completely surrounded with oil. I am using a cartidge heating element slipped inside the dry pipe. The heater is rated for 225 watts at 220 volts, I am powering this with 110 volts so it is making 60 some watts. By underating this heater it is surviving just fine even though it is not installed with heat transfer grease or anything else. 60 watts is proving to work fine to keep the oil warm, maybe 80 deg f, no thermostat needed.

The oil output is in the center of the rounded bottom, I installed a 3 1/2 inch standpipe up inside the tank, this is made from 1/4 inch pipe, it has a wire tank strainer screwed on the inside end to catch any big chunks, hopefully the 3 1/2 inches of space below the opening of the standpipe will collect most of the crud that settles out of the oil and keeps it out of the pump.

This tank will supply enough fuel for from 4 to 10 days of heat depending on the temps, should not have to drag fuel cans through a blizzerd this winter, I hope.

upright_pic_of_1st_st_fuel_tank.JPG (50 Kb, 92 downloads)

This shows the strainer basket that sets just inside the top of the 25 gallon fuel supply tank, the bottom is made from perferated sheetmetal with 1/16 inch holes, I also place a layer of window screen on the inside, just another step in trying to keep chunks out of the pump gears.

removable_strainer_-_resized_and_tweeked.JPG (88 Kb, 50 downloads)

I machined up a version of the Sanders shallow 12 degree burner bowl just as a test. INTERESTING -- Once I got it hot enough to burn it performed almost exactly as describe in the Sanders heater article. The oil pool did indeed expand and contract in diameter as the fuel feed was changed, the oil pool did stay liquid while burning, the pool of oil boiled and stayed moist all during the 8 hour test. I never could get the air adjusted for this burner.

Sanders stated he had to use a propane torch to get his burner hot enough to burn veg, likely so, I made this burner to set inside my SS mixing bowl burner, this allowed me to easily adjust the position of the bowl in its canister to get the Sanders burner plate exactly level, the bowl also would catch any un-burnt fuel that dripped off the burner plate. The bowl also made startup easy by holding the startup fuel and placing the flame under the Sanders plate for max heatup.

Well -- I poured about a half-cup of diesel into the bowl, lit it, turned on the oil drip, it took almost an hour before the sanders plate got hot enough to ignite the fuel in it. Once it was warmed up and burning it performed as Sanders explained.

BUT - or WHEN - or HHMMM -- One thing not addressed in the Sanders article is SMOKE, the Sanders burner did not put out as much heat as the SS bowl, it also puts out noticably MORE smoke than the SS bowl. Thinking about this a bit I suspect the same properties that keep the crud on the Sanders burner plate soft are the same properties that make it smoke, namely, the plate does not get hot enough to vaporize the liquid fuel completely, this somewhat poor combustion leads to less heat and more smoke. The fuel stays as a liquid pool in this burner, the plate gets hot enough to boil the fuel pool but not hot enough to boil it away, this liquid keeps the plate surface moist so the amount of hard coal is minimized. Unfortunatly, the slowly boiling fuel is not producing as fine of a vapor as the very hot SS bowl.

With the bowl burner I don't have any liquid fuel in the bottom after just a few minutes from startup. Since the bowl is located partially inside another closed container the base of the bowl stays extremly hot, this vaporizes the fuel drops virtually instantly into vapor rather than fine droplets, the flame is also down inside the sloping sides of the bowl so it imparts a lot of heat to the bowl. Since this bowl burner does not create nearly as much coal as a turk burner I have some suspicion that the fuel droplets may actually be partially vaporizing in the air even before they hit the hot bottom of the bowl, don't know for sure, but I don't hear every drop sizzle as the burner runs ? Sizzle is not exactly a correct discription either, if the oil is free of water I don't hear any sound from each droplet of fuel, it is only if the fuel is a bit wet that I hear an almost instantanious SSSTT when the drop hits the bowl. I may try reducing the size of the dropletes, smaller drops should vaporize even faster, hopefully in the air. The fuel now exits through an 1/8 inch ID steel line, I will braze this opening over and drill a 1/16 inch hole as a test (eventually, likely next summers change).

The Sanders article did mention that the fuel droplets needed to be tiny or they would splash when they hit the liquid fuel pool, I had this problem, a lot of fuel got splashed out against the inner wall of the burn chamber, this oil mixed with soot and made a greasy black mess as it slowly ran down the inner wall, also cought fire once I switched back to the bowl burner, had to shut down all the draft controls as this splashed oil, and all the soot on the wall of the burner and in the flue pipe, lit off, LOTS of heat for a few minutes.

burner_plate_in_bowl.JPG (48 Kb, 92 downloads)

I am now controlling the wiper motors with actual pulse width modulated DC motor speed controls, this seems to have eliminated the slow-speed stall problem I was having last winter when using the analog voltage created by using Veriac-style veriable transformers powering the battery chargers. Pulsed 12 volts creates a lot more torque from the motor than an actual extremely low voltage.

I don't know just what vehicles my two wiper motors came from, something GM from the 80's I think, They must have different gear ratios as one will run down to as slow as 7 seconds/rev, the other will run down to as slow as 17 sec/rev? They both run up to around 3 rev/sec when powered directly with 12 volts, plenty fast for the J model pumps and these burners. If using "A" model pumps, especially the one rated at 3450 RPM, it may actually take a faster turn speed to move the required amount of oil. Luckily there are all sorts of 12 volt gearmotors available at surplus prices, just have to find the best gear ratio.

Most of the wiper motors I see now in the surplus catalogs are from modern small cars, these have two speeds and seem to turn faster than the old ones I am using, they are also considerably smaller in size. I suspect the 1725 RPM rated "A" pump will move enough oil to turn it with a wiper motor, they move 1.6 times as much oil per rev as the 3450 RPM rated pump.

MY wiper motors start turning at about 1.5 amps and draw about 4 amps at full speed. I am using cheap 6 amp automotive battery chargers connected to 12 amp-hr gell cell batteries to power these. The charger keeps the battery charged and powers the motor as long as there is normal 110 volt power, the battery will take over and last for anywhere from 3 to 8 hours if there is a power outage. Most of the local power outages happen during the night but only last for a few hours, hopefully I won't even wake up.

I will be experimenting with incorperating an optical flame sensing device to shut the power down if the flame goes out. I had flameouts happen a couple times last winter during high winds, of course, it was during the middle of the night. The fuel kept dripping but I had a catch container under the stove so no fuel got completely away, just a bit messy cleaning everything up after the fuel had strained through the layer of soot in the bottom of the stove.

great to get this discussion moving again..I will dust off my project and get rolling too..just added 20x40 to my shop so I'll need it..Tigman

Hey Tim,

I have started working on my heater again of course I waited till the last minute as usual. I finally had my 24x36 shop built and installed my heater. I see that draft is very critical and being that as so I have installed many furnaces in my time (Im a HVAC tech) and I remebered a device called a barometric damper. I did a search on this forum and didnt notice not one mention of it. Has any one ever tried one? This may be a answer to alot of problems I will be picking one up in the morning to try it. here is a link that gives a good definition http://www.bacharach-training.com/combustionzone/draft1.htm
But in short it can be adjusted to maintain a certain draft across a furnace. What do you think?

HHMMM -- reading through the link, it looks like these controls are plumbed in paralell with the firebox to automatically allow additional room air to enter the flue pipe when the chimney creates more suction, like when high wind blow across the top of it. may help, don't know realy. I am not sure just what mechanism causes the rare flame blowout in this particular stove, it is caused during high winds so I assume the extra suction created in the chimney causes a lot more air to be drawn down the combustion intake air tube, this air blows directly into the burner bowl from above so this is likely what blows the fire out, adding a swinging damper to the flue stack may work by allowing the excess suction to swing the damper open a bit reducing the suction, It would be worth a try.

I run both a manual flue damper and a manual combustion air intake damper. I open both about 1/4 open and get the burner going, wait til everything is up to temp (maybe 10 minutes), then close the exhaust flue damper just til I get smoke coming from the verious cracks in the stove, usually get a bit of smoke out of the combustion intake air tube then also, open the exhaust flue damper just enough that the smoke goes away, this keeps a LOT more heat inside the stove rather than letting it go up the chimney. My exhaust flue dampers are almost always 2/3 to 3/4 closed. I then open up the combustion air intake damper while watching the flame through a 3/8 inch hole drilled through the side of the burner tank (the burner door is closed). This small viewing hole lets me see how the fire is burning under it's normal draft conditions, as long as the smoke is going up the chimney air is moving inward through this small hole so no flames excape, the air flow also keeps it clear of soot. I open the intake air damper til the flame becomes a bit erratic showing big flickering changes in the brightness of the flickering flame, then I close the damper down a bit till the flame becomes almost steady and uniform (there will always be a bit of flickering).

I sort of dought there is enough draft going through this stove to let the added barometric damper to adjust much for the excess draft unless there was a BIG change in the chimney suction, this may be too late to keep from blowing out the small flame that I normally run in this stove, don't know, it would be an interesting test. If I find one of these units chaep I will give it a try. I will install this automatic damper on the chimney side of my manual flue damper, the manual flue damper is always pretty well closed down so the automatic damped should help control the max suction being created from the chimney, the almost-closed manual damper may keep the smaller, somewhat automatically controled, chimney suction from being felt so much on the firebox side of it.

Just had a thought -- I would want to be able to manually close the automatic damper in case of a soot fire in the flue pipe or chimney, I have had this happen a couple times during early testing of this heater, shutting down all the dampers controlled it to a low heat smolder that cleaned out all the soot nicely over a few minutes, opening any damper a bit would easily cause the burn to rise to a point of turning the tin flue pipe cherry red just in front of the manual flu damper.

I have been having EXTREMELY GOOD SUCCESS with very tiny flames so far this year. I suspect this has to do with much steadier fuel control as this is the only area that I have made improvements in. The mods to the Suntec pump allow much better fuel flow to the internal gears and the pulse-width modulated wiper motor speed control has allowed extremely precise fuel flows even at very low pump speed settings. The temps have been pretty warm for the last several weeks, high 50's to low 60's f during the day, mid 40's during the night. This is cool enough that I needed a bit of heat but not much. I have been running th pump at around 6-7 seconds per rev, this puts one drop from the 1/8 inch drip tube opening into the burner bowl every 1 to 1 3/4 seconds. I could not get this small of a burn last year without flameouts every few hours, I have only had one flamout so far this year over the 3-4 weeks I have been running the stove almost contiuously.

At this low fuel rate there is almost NO noticable smoke from the chimney, have to stop and stare to see it, and the coal in the burner bowl is maybe a 1/8 to 3/16 inch thick in a 2 inch diameter circle directly where the fuel drops land, after burning continuously for 3 days. I am a bit amazed with this, way better than I had ever expectid.

The one flameout that I had was when I tried burning the bowl burner for 4 days without cleaning, this was too long to go without cleaning, the coal had formed sort of a "volcano cone" around the area where the fuel drops were hitting the bowl, I suspect this was inhibiting the combustion air from mixing with the fuel vapor as it should.

Somewhat surprisingly, even at this low combustion air volume the dry "leafletts" of soot were being blown out of the bowl and deposited loosly on the inside bottom of the stove. There was maybe a 2 inch deep layer of this dry pillowy soot on the bottom of the stove after a couple weeks of this low burn. The soot was easily removed using a broad putty knife as a scoop.

The precise fuel control is also pretty repeatable. I have a thermometer mounted to the outside of the burner tank, just where the flue pipe mounts to the top of the tank. At the 7 second/rev fuel feed the thermometer reads 130 deg f, at the 6 sec/rev setting the thermometer reads 150 deg f, 4 sec/rev reads 200 deg f, this is with a 10 inch tabeltop type fan, set on it's lowest speed, and blowing across the flue and burner tank from about 6 feet away. If the fan is turned off the thermometer reads from 30 to 50 degrees higher. These temps have been repeatable every time I have adjusted the pump speed.

i had another flameout the other night and ran about a gallon of oil through the soot inside the stove, no big problem as the oil drips out into a catch bucket under the stove but it takes about an hour to crank up the burner and burn all the oil inside the soot off at high heat. The stove gets up to around 400 degrees during one of these clean-it-out burns so I watch it closely, this all takes about an hour and is a bit of a nuisance.

This has prompted me to finally install a flame detection device that turns off the pump motor once the flame goes out. This has worked fine during testing but I have only had it in operation but a couple days and have not yet had another flamout, but I expect it to work just fine.

A quicky schematic is attached below. This is a pretty simple 3 transistor circuit plus a cheap small infrared optotransistor used as the sensor. I built this to tape to a normal automotive accessory relay but it turned out to be a bit bigger than I had envisioned, still stuck it to the ralay but it is a bit lumpier than intended.

The idea is based on a capacitor that slowly charges up over 15 seconds, every time there is a flicker of flame being seen by the sensor the capacitor gets discharged and has to start charging again from almost zero volts, if no flame is seen for the entire 15 seconds the capacitor charges up to just under 12 volts, once this point is reached the output section of the circuit turns off the relay, this switches off the 12 volt power to the motor speed control unit.

In the drawing I analagized these 3 sections to a bucket being filled through a small hose, the flame sensor repeatidly empties the bucket, and the output circuit acts like a float switch that turns everything off when the bucket gets full.

sensor_circuit_schematic.JPG (67 Kb, 55 downloads)

This a picture of the little circuit board mounted on top of the relay.

2_-_board_with_added_power_filter_cap.JPG (31 Kb, 43 downloads)

Side view of the board and relay.

8_-_board_with_added_power_filter_cap._-_front__JPG.JPG (22 Kb, 29 downloads)

The way the sensor watches the flame is through a small hole drilled through the burner tank wall, this hole is located about a half inch above the burner bowl. The sensor is a tiny plastic device that is the same size as the smaller sized T1 led's. Since it is plastic I wanted to keep it as cool as possible so it is mounted in the far end of a one inch length of 1/8 ID, 5/32 OD, brass hobby tubing. The tube gets the sensor back away fron the hot stove and also limits it's field of view to the light coming directly in to the very nerrow tube, this helps eliminate false triggering that could be caused by stray light in the room rather than light coming directly from the flame.

To further remove the sensor from heat the brass tube is press-fit through a piece of phenolic, this phenolic also acts as the mount to the stove. The phenolic is held to the stove by a 1/4 inch bolt, wing nut, and spacer. These hold the phenolic about 1 1/2 inches away from the side of the stove burner tank. the wing nut allows the assembly to be swung sideways, or removed, to allow the viewing hole to be cleaned if nescisary ( cool air flows INWARD through this hole when the burner is burning, this tend to keep it free of soot).

This first-cut sensor mount design did not quite work as planned due to everything still getting too hot, The modification will be shown in the next few posts.

This picture shows the basic mount and one of the sensors. The actual sensor is located inside the rear end of the brass tube that is covered with shrink tubing.

5_-_T1_IR_sensor_next_to_mount.JPG (36 Kb, 33 downloads)

This is a closer view of the sensor and the sensor mounting tube.

4_-_T1_IR_sensor_next_to_mount.JPG (29 Kb, 28 downloads)

This shows the two holes drilled throught the burner tank. The top hole is tapped 1/4 -20 for the mounting stud, the lower hole is the viewing hole. I drilled a 1/4 inch hole for the viewing hole, turned out to be bigger than needed, a 3/16, or probably evan 1/8 inch, hole would be plenty, the flame image on the face of the phenolic sensor mount from the 1/4 inch hole is actually 1/2 to 5/8 inch in diameter, plus being inverted, the hole acts like the pinhole in a pinhole camera.

sensor_holes_in_burner_tank.JPG (16 Kb, 27 downloads)

This shows the sensor mounting stud screwed into the top hole.

sensor_mounting_stud_on_burner_tank.JPG (13 Kb, 24 downloads)

This shows the entire original sensor assembly mounted to the burner tank.

3_-_sensor_mounter_on_burner_tank.JPG (16 Kb, 33 downloads)

This shows the updated sensor assembly with an additional thin aluminum heat shield installed. This heat shield has a 3/16 hole drilled in line with the sensor tube, this keeps any excess heat from reaching the phenolic. I did a stray-light test using a flashlight and found one tiny angle from the back of the sensor where enough light was reflected off the back of the shiny heat shield to trigger the sensor, I will paint the back of the heat shield with flat black paint in the future.

The heat shield reflects most of the light/heat back to the burner tank, the shield stays cool enough to touch and the reflected light make a pretty dancing 1 1/2 inch diameter rosette of orange light centered around the flame viewing hole.

The flame sensor tube has also been moved upwards about a quarter inch to locate it closer to the brighters portion of the flame image, this is due to the flame image being inverted, didn't consider the inverted image in my original thinking.

Without the heat shield installed the dark phenolic got too hot to touch, and the sensor tube was also way hotter than I wanted, with the heat shield installed, the phenolic stays cool enough that you never have to pull your finger off it, the sensor tube stays at just slightly warmer than room ambient.

2_-_sensor_mounter_on_burner_tank.JPG (12 Kb, 35 downloads)

Side view of shielded sensor mount.

8_-_sensor_mount_with_added_heat_shield_-_side__.JPG (15 Kb, 22 downloads)

Front view of shielded sensor mount.

1_-_sensor_mount_with_added_heat_shield_-_front.JPG (18 Kb, 30 downloads)

having trouble reading schematic..but I assume its like a bosch relay circut for turning lights, turned on by turn signal flash? Tigman